Abstract
Objectives
Our study aimed to describe users of a free online contraception service, compare online emergency contraceptive pill (ECP) users with online oral contraceptive (OC) users, and describe patterns of use of online ECP and OC over time, including transition from ECP to more effective forms of contraception.
Study design
Analysis of routinely collected, anonymized data from a large, publicly funded, online contraceptive service in the United Kingdom from April 1, 2019 to October 31, 2021.
Results
The online service provided 77,447 prescriptions during the study period. About 84% for OC and 16% for ECP, of which 89% were ulipristal acetate. ECP users were younger, lived in more deprived areas, and were less likely to be of white ethnicity than OC users. About 53% ordered OC only, but 37% ordered both ECP and OC. Among those prescribed both OC and ECP (n = 1306), 40% had one method as dominant, 25% appeared to move from one method to the other (11% ECP to OC, 14% OC to ECP), and 35% continued to use both methods.
Conclusions
Online services are accessible to diverse young populations. While the majority of users order OC only, our study suggests that where there is free, online access to both OC and ECP, and those ordering ECP are always offered free OC, transition to more effective ongoing forms of contraception is uncommon. Further research is needed to understand whether online access to ECP increases its attractiveness and reduces likelihood of transition to OC.
Implications
This study demonstrates that free, online contraceptive services are accessible to ethnically and socioeconomically diverse users. It identifies a subgroup of contraceptive users who combine use of OC and ECPs over time, and suggests that improved access to ECP may alter contraceptive choices.
Keywords: COVID-19, Digital, Emergency contraception, Online, Oral contraception, Telehealth
1. Introduction
Emergency contraception can be used to prevent pregnancy after unprotected sex, through insertion of a copper intrauterine device or emergency contraceptive pills (ECPs) [1]. In Great Britain, 7% of the young women use emergency contraception each year [2], mainly ECP [3]. Timely access to emergency contraception maximizes effectiveness [4] and previous research has measured access from pharmacies, family doctors (general practitioners), and sexual health clinics [5], [6].
ECP provision offers opportunities to support transition to a more effective method. In one UK study, 56% of the women accessing ECP face-to-face from a specialist contraceptive service chose to continue or begin more effective ongoing hormonal contraception immediately [7]. Previous research has documented a pattern of “switching” to a more effective method of contraception over time [5]. Providing the progestogen-only pill at the time of emergency contraception access from pharmacies increased subsequent use of effective contraception by 20% (95% CI: 5.2–35.0) [8].
Online contraceptive services may improve access [9] and online provision increased rapidly during the COVID-19 pandemic as face-to-face services were closed due to staff redeployment and to control transmission. Through online services, combined hormonal methods, progestogen-only pills, ECP, and self-injectables are accessible by mobile phone, at any time and without the perceived or real stigma associated with visiting services [10]. However, digital services may not increase access for everyone equally and may disadvantage specific groups and widen health inequalities [9], [11]. Online ECP provision could prompt uptake of more effective contraception through information to support contraceptive choice or prompts to make an immediate order for oral contraceptives (OCs) or self-injectables [9]. However, digital services do not offer the opportunities for immediate provision of long-acting reversible contraceptives (LARCs) available through face-to-face care.
SH:24 is one of a number of online sexual health providers in the United Kingdom [12]. It offers ECP, OC (combined hormonal contraceptive and progestogen-only pills), and self-injectables to people aged over 16, by post after an online medical history supported by clinical conversations by SMS and telephone. Ongoing prescriptions for OC are offered lasting 3, 6, or 12 months.
The service offers ulipristal acetate (UPA, 30 mg) as first line and levonorgestrel (LNG 1500 µg) when UPA is contraindicated (e.g., cases of severe asthma controlled by oral glucocorticoids, breastfeeding) [13]. Users input time of latest pregnancy risk and the service calculates the time the medication will take to reach the service user by post (24 hours tracked delivery), with an automatic redirect to face-to-face care those who could not receive it on time. SH:24 provides information on all methods of contraception, including a contraception comparator tool [14] and prompts users to order ongoing contraception ( Fig. 1). Users complete a digital form that collects clinically relevant information that is assessed by a specialist clinician. This includes information on allergies, interacting medication, or contraindications to UPA or LNG following national guidance issued by the UK Faculty of Sexual and Reproductive Health Care [13]. All ECP orders include an information card that provides information about interactions between ECP and OC.
Fig. 1.
Automated text-message prompt sent to all those ordering ECP from SH:24's online contraceptive service. ECP = emergency contraception pill.
We were interested in the impact of online emergency contraceptives on access, patterns of ECP use, and transition to more effective ongoing methods of contraception. Our study aimed to describe users of a free online ECP service, compare online ECP with online OC users, and describe patterns of use of online ECP and OC over time, including transition from ECP to more effective forms of contraception.
2. Methods
We analyzed anonymized, routinely collected data from users of a publicly funded, online sexual and reproductive health service (SH:24) [12], operating across 51 of the 217 upper tier local authority areas in the United Kingdom [15]. We included all orders placed between April 1, 2019 (when the service began) and October 31, 2021; 94% of the orders occurred after April 2020. Since 99% of the orders for ongoing contraception are for OC and orders for the patch, vaginal ring, and self-injectable methods are low, we only included OC pills (combined hormonal contraceptives and progestogen-only pills) in this analysis.
Our variables of interest were contraceptive ordered (coded as ECP or OC), age, a nationally assigned index of deprivation calculated by area of residence (neighborhood Index of Multiple Deprivation [16]), sexuality, and ethnicity (self-reported using national census categories [17]).
We performed regression analysis using both the full list of ethnic groups and a “condensed” list (e.g., black African, black Caribbean, and black other condensed into a single category “black”). We did not find significant differences within the “condensed” categories; so for ease of interpretation, we present the “condensed” regression here.
We categorized users into two groups: OC users (anyone who had received OC) and ECP users (anyone who had received ECP). Some users received both OC and ECP. For users who ordered both OC and ECP, we calculated the time between first order of each contraceptive type. We compared the groups using logistic regression in STATA14. The outcome variable was type of contraceptive used. We included age, ethnicity, and deprivation as explanatory variables, as variability in contraceptive use within these categories had been documented in the published literature [6], [11], [18]. We did not include sexuality as a variable in the model as it did not vary between the two outcome groups ( Table 1).
Table 1.
Age, ethnicity, sexuality, and deprivation of SH:24's online oral contraceptive and emergency contraceptive pill users (United Kingdom, between April 2019 and October 2021)
| ECP users N (%) | OC users N (%) | |
|---|---|---|
| Total | 6021 (100) | 36,032 (100) |
| Age | ||
| 16–20 | 1979 (33) | 10,430 (29) |
| 21–25 | 2005 (33) | 11,100 (31) |
| 26–30 | 1087 (18) | 6475 (18) |
| 31–35 | 553 (9) | 3513 (10) |
| 36–40 | 247 (4) | 2173 (6) |
| 41–45 | 106 (2) | 1283 (4) |
| 46 and over | 44 (1) | 1058 (3) |
| Ethnicity | ||
| White British | 4209 (70) | 23,135 (64) |
| White other | 303 (5) | 1865 (5) |
| Black | 572 (10) | 1345 (4) |
| Asian | 329 (5) | 1218 (3) |
| Chinese | 24 (<1) | 181 (<1) |
| Mixed | 438 (7) | 1469 (4) |
| Other | 40 (1) | 203 (1) |
| Not recorded | 106 (2) | 6616 (18) |
| Sexuality | ||
| Heterosexual | 5342 (89) | 26,452 (89) |
| Homosexual | 18 (<1) | 91 (<1) |
| Bisexual | 391 (7) | 2009 (7) |
| Not recorded | 270 (5) | 1106 (4) |
| Deprivation quintile | ||
| 1 (most deprived) | 1721 (29) | 7280 (20) |
| 2 | 1318 (22) | 5749 (16) |
| 3 | 1058 (18) | 4768 (13) |
| 4 | 710 (12) | 4036 (11) |
| 5 (least deprived) | 652 (11) | 4356 (12) |
| Not recorded | 562 (9) | 9843 (27) |
ECP, emergency contraception pill; OC, oral contraceptive.
In addition, we generated user “journeys” as sequences of contraceptive orders. We used R [19], coding each order as ECP (e), OC (o), or a combined order (x), and then concatenating the code letters in date order for each user. We manually classified the journeys into one of seven “journey types” ( Fig. 2).
Fig. 2.
Number of users of SH:24's online contraceptive service, by number of orders and product ordered (United Kingdom, between April 2019 and October 2021).
We excluded users who placed their first order after April 30, 2022; so all users had at least 6 months of orders to analyze. We excluded orders if they were placed but not prescribed (e.g., because there was a clinical contraindication). Our analysis of user journeys only included those who had placed enough orders to detect patterns of use (four or more).
3. Results
The online service provided 77,447 prescriptions during the study period. About 84% were for OC and 16% for ECP. Progestogen-only pills made up 57% of the OC prescriptions, and combined pills 43%. UPA made up 89% of the ECP prescriptions, with the remainder for LNG.
We found that users from more deprived areas were overrepresented in both groups of contraceptive users (Table 1). ECP users were on average younger, living in more deprived areas, and less likely to be of white ethnicity than OC users ( Table 2).
Table 2.
Odds of being an emergency contraceptive pill user versus an oral contraceptive user for each demographic characteristic, among users of SH:24's online contraceptive service (United Kingdom, between April 2019 and October 2021)
| OR | Lower 95% CI | Upper 95% CI | p Value | |
|---|---|---|---|---|
| Age | ||||
| 16–20 | 1 | |||
| 21–25 | 0.89 | 0.83 | 0.95 | 0.001 |
| 26–30 | 0.82 | 0.75 | 0.89 | <0.001 |
| 31–35 | 0.78 | 0.70 | 0.86 | <0.001 |
| 36–40 | 0.58 | 0.50 | 0.66 | <0.001 |
| 41–45 | 0.39 | 0.31 | 0.47 | <0.001 |
| 46 and over | 0.19 | 0.14 | 0.26 | <0.001 |
| Ethnicity | ||||
| White British | 1 | |||
| White other | 0.91 | 0.81 | 1.04 | 0.162 |
| Black | 2.62 | 2.35 | 2.91 | <0.001 |
| Asian | 1.51 | 1.32 | 1.71 | <0.001 |
| Chinese | 0.70 | 0.46 | 1.08 | 0.105 |
| Mixed | 1.77 | 1.58 | 1.98 | <0.001 |
| Other | 1.23 | 0.87 | 1.73 | 0.241 |
| Not recorded | 0.09 | 0.08 | 0.11 | <0.001 |
| Deprivation quintile | ||||
| 1 (most deprived) | 1 | |||
| 2 | 0.95 | 0.87 | 1.03 | 0.187 |
| 3 | 0.94 | 0.86 | 1.02 | 0.131 |
| 4 | 0.72 | 0.65 | 0.79 | <0.001 |
| 5 (least deprived) | 0.59 | 0.53 | 0.66 | <0.001 |
Data on deprivation and ethnicity were incomplete (26% and 17% missing, respectively).
We categorized each user who had received four or more contraceptive products (n = 3547) into one of seven types depending on the pattern of orders (Fig. 2).
The majority of users who made at least four orders from the service (n = 1896, 53%) only ordered OC from SH:24, and a further 345 (10%) only ordered ECP.
The rest of the contraceptive users (n = 1306, 37%) ordered both OC and ECP during the study period. Within this group, 40% appeared to have largely settled contraceptive strategies (“Broadly OC” or “Broadly ECP”), and 25% changed contraceptive method during the study period (139 (11%) moved from ECP to OC, and 178 (14%) from OC to ECP) ( Table 3). The remaining 461 (35%) were categorized as “mixed”; they ordered both ECP and OC over time, not showing a strong preference for any one method nor appearing to transition from one method to another. Of those who used both ECP and OC, 7.6% ordered a different method from their previously ordered method four or more times and the highest number of such changes was eight over 19 months.
Table 3.
Summary of user journeys among users of SH:24's online contraceptive service (United Kingdom, between April 2019 and October 2021) (includes only those who ordered from the service four or more times)
| Category and definition | Frequency | Percentage (of all orders) | Percentage (of orders that include ECP and OC) | Mean age (y) | Mean deprivation index scorea | % White British |
|---|---|---|---|---|---|---|
| Only ECP: user has only ordered ECP from SH:24 | 345 | 10 | N/A | 23.5 | 27.1 | 67 |
| Broadly ECP: all orders are ECP except 1, or except 2 if >8 orders, or except 3 if >12 orders | 326 | 9 | 25 | 23.4 | 29.8 | 69 |
| OC→ECP: starts with at least 1 order of OC, ends with at least 2 orders of ECP | 178 | 5 | 14 | 23.9 | 27.7 | 75 |
| Mixed: pattern fluctuates, no clear pattern can be discerned, does not fall into any of the other categories | 461 | 13 | 35 | 23.3 | 27.7 | 74 |
| ECP→OC: starts with at least 1 order of ECP, ends with at least 2 orders of OC | 139 | 4 | 11 | 22.8 | 28.3 | 73 |
| Broadly OC: all orders are OC except 1, or except 2 if >8 orders, or except 3 if >12 orders | 202 | 6 | 15 | 22.7 | 25.9 | 77 |
| Only OC: user has only ordered OC from SH:24 | 1896 | 53 | N/A | 25.2 | 23.6 | 83 |
ECP, emergency contraception pill; OC, oral contraceptive.
For deprivation index, higher score = more deprived.
The characteristics of members of each group are summarized in Table 3. Users who had only ever ordered OC tended to be older, and were more likely to be white, and to live in less-deprived neighborhoods. Those in other journey types were more similar to each other in terms of age and deprivation, although % white British and reliance on ECP appear to be inversely correlated.
Finally, we considered the time interval between first ECP and first OC orders for women who had received both. The most common time interval was 0 days, that is, ordered on the same day, suggesting an intention to switch to an ongoing, more effective method. However, the distribution of intervals is approximately symmetrical, with equal numbers of users ordering ECP then OC as ordered OC then ECP, suggesting no overall trend from ECP to OC. The event of requiring ECP is not the only time at which users choose to change contraceptive methods—almost as many (558 vs. 608) order their first OC within 12 weeks of their first ECP as order OC on the same days as first ECP.
4. Discussion
This study describes patterns of transition between OC and ECP within an online service during the COVID-19 pandemic, when access to alternatives was limited. Our longitudinal data are unusual and enable description of patterns of online contraception use over time. These can be difficult to study as records of face-to-face contraceptive providers are rarely linked. Previous research has used infrequent surveys with a risk of recall bias [5], [20], [21], or linked records over shorter time periods [8].
We found that online contraceptive users, and particularly those who order ECP, were young, ethnically diverse, and disproportionately from more deprived neighborhoods. Our data on deprivation and ethnicity support findings during the pandemic [22], suggesting that online services are accessible to diverse populations of young people and reach groups previously identified as experiencing barriers to access [6], [11]. Our data on age of online ECP users are comparable with national statistics on age of face-to-face ECP users in 2019/2020 [18].
Although most (63%) online users in this study pursued a consistent contraceptive strategy (usually OC), a third of users ordered both OC and ECP. While 11% of this group made the transition from ECP to consistent OC use, 35% had an unpredictable pattern of use that may have included both simultaneous (e.g., after a missed OC pill) and sequential (i.e., a change of contraceptive method) use and 14% appeared to transition away from more effective ongoing OCs to using ECPs.
Rates of transition to more effective ongoing methods of contraception in this study are lower than previously reported. For example, self-reported use of effective contraception at 4 months after ECP use was 41% in the control arm of a recent trial aiming to increase uptake of effective contraception through the provision of OC as standard (without the need to request it) when emergency contraception is obtained from a pharmacy [8]. Our findings are not directly comparable with that study as it used self-reported data and different follow-up processes, study populations, and means of contraceptive access. However, the difference suggests that even when users are able to use an online service for ECP, and are prompted to order OC at the same time, barriers to transition remain.
The online service studied is unusual because OC and ECP are equally accessible, 24 hours a day, from home (assuming no clinical contraindications). The service provides a novel context in which to study contraceptive choices. The frequency of method change identified is higher than found in studies with face-to-face access, suggesting a more complex contraceptive trajectory [5]. Improved access to ECP may increase the attractiveness of ECP as a regular contraceptive strategy or reduce motivation to switch to more effective forms. Alternatively, these complex patterns of use may have been underreported in previous studies.
The limitations of this study include that online users may also have accessed contraception from other sources during the study period [23], and that we used delivery addresses to infer deprivation but home addresses may have been different. Data on ethnicity and deprivation were incomplete for 17% and 26% of the customers, respectively. We linked each order to a “user” via mobile phone number and if this number changed ownership, or users clandestinely ordered for others, “users” may represent more than one person. We did not analyze the type of OC, and simultaneous or sequential use of different methods. Users of OC only may be underrepresented compared with users of ECP, as prescriptions can last up to 12 months, making them less likely to reach the threshold of four orders. However, across all journey types, more than 60% of the OC prescriptions were for 3 months only, with no significant differences in prescription length between users with different journey types.
This study demonstrates that these online services were accessible to young people of diverse ethnic backgrounds. Most online service users order OC consistently, but among those who order both OC and ECP, only 11% transition from ECP to OC and many adopt complex patterns of use of both methods over time.
Footnotes
Conflicts of interest: P.B. is the medical director of SH:24. M.T. and V.I. are employed by SH:24. F.B. undertook this work while on a training placement at SH:24. The authors report no conflicts of interest.
Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
References
- 1.World Health Organisation. Selected practice recommendations for contraceptive use. 3rd ed.; 2016. 〈https://www.who.int/publications/i/item/9789241565400〉 (cited December 23, 2022).
- 2.Black K.I., Geary R., French R., Leefe N., Mercer C.H., Glasier A., et al. Trends in the use of emergency contraception in Britain: evidence from the second and third National Surveys of Sexual Attitudes and Lifestyles. BJOG. 2016;123(10):1600–1607. doi: 10.1111/1471-0528.14131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Johnson A. National Survey of Sexual Attitudes and Lifestyles, 2010–2012 [data collection]. 2nd ed.: UK Data Service. SN: 7799; 2021. 〈https://www.natsal.ac.uk/natsal-survey/natsal-3〉 (cited December 23, 2022).
- 4.Haeger K.O., Lamme J., Cleland K. State of emergency contraception in the U.S., 2018. Contracept Reprod Med. 2018;3(20) doi: 10.1186/s40834-018-0067-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wellings K., Brima N., Sadler K., Copas A.J., McDaid L., Mercer C.H., et al. Stopping and switching contraceptive methods: findings from Contessa, a prospective longitudinal study of women of reproductive age in England. Contraception. 2015;91(1):57–66. doi: 10.1016/j.contraception.2014.09.008. [DOI] [PubMed] [Google Scholar]
- 6.Cameron S.T., Gordon R., Glasier A. The effect on use of making emergency contraception available free of charge. Contraception. 2012;86(4):366–369. doi: 10.1016/j.contraception.2012.02.018. [DOI] [PubMed] [Google Scholar]
- 7.Baird A.S., Trussell J., Webb A. Use of ulipristal acetate and levonorgestrel for emergency contraception: a follow-up study. J Fam Plan Reprod Health Care. 2015;41(2):116–121. doi: 10.1136/jfprhc-2013-100733. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cameron S.T., Glasier A., Mcdaid L., Radley A., Baraitser P., Stephenson J., et al. Use of effective contraception following provision of the progestogen-only pill for women presenting to community pharmacies for emergency contraception (Bridge-It): a pragmatic cluster-randomised crossover trial. Lancet. 2020;396:1584–1594. doi: 10.1016/S0140-6736(20)31785-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rezel-Potts E., Free C., Syred J., Baraitser P. Expanding choice through online contraception: a theory of change to inform service development and evaluation. BMJ Sex Reprod Health. 2020;46(2):108–115. doi: 10.1136/bmjsrh-2018-200297. [DOI] [PubMed] [Google Scholar]
- 10.Free C., Lee R.M., Ogden J. Young women’s accounts of factors influencing their use and non-use of emergency contraception: in-depth interview study. BMJ. 2002;325(7377):1393. doi: 10.1136/bmj.325.7377.1393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Poole L., Ramasawmy M., Banerjee A. vol. 6. Elsevier Ltd.; 2021. Digital first during the COVID-19 pandemic: does ethnicity matter? pp. e628–e630. (Lancet Public Health). [DOI] [PubMed] [Google Scholar]
- 12.SH:24. SH:24 website. 〈https://sh24.org.uk/〉 (cited December 23, 2022).
- 13.SH:24. SH:24 contraception comparison page. 〈https://sh24.org.uk/contraception/compare?〉 (cited December 23, 2022).
- 14.Faculty of Sexual and Reproductive Healthcare. FRSH guideline: emergency contraception; 2020. 〈https://www.fsrh.org/documents/ceu-clinical-guidance-emergency-contraception-march-2017/〉 (cited December 23, 2022).
- 15.UK Government. Understand how your council works (GOV.UK); 2022. 〈https://www.gov.uk/understand-how-your-council-works〉 (cited December 23, 2022).
- 16.Ministry of Housing C and LG. English indices of deprivation 2019; 2019. 〈https://www.gov.uk/government/statistics/english-indices-of-deprivation-2019〉 (cited December 23, 2022).
- 17.Office for National Statistics. Ethnic group, national identity and religion. 〈https://www.ons.gov.uk/methodology/classificationsandstandards/measuringequality/ethnicgroupnationalidentityandreligion〉 (cited December 23, 2022).
- 18.NHS Digital. Statistics on sexual and reproductive health services (contraception): data tables. 〈https://digital.nhs.uk/data-and-information/publications/statistical/sexual-and-reproductive-health-services/2019-20/data-tables〉 (cited December 23, 2022).
- 19.R Core Team . R Foundation for Statistical Computing; Vienna: 2021. R: a language and environment for statistical computing.〈https://www.R-project.org/〉 [Google Scholar]
- 20.Jones R.K., Tapales A., Lindberg L.D., Frost J. Using longitudinal data to understand changes in consistent contraceptive use. Perspect Sex Reprod Health. 2015;47(3):131–139. doi: 10.1363/47E4615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lindh I., Andersson Ellströ A., Blohm F., Milsom I. A longitudinal study of contraception and pregnancies in the same women followed for a quarter of a century. Hum Reprod. 2010;25(6):1415–1422. doi: 10.1093/humrep/deq095. [DOI] [PubMed] [Google Scholar]
- 22.Goward C., Clune M., Ebbutt K., Perera S. P53 an analysis of service users accessing a routine contraception service via an online platform during the COVID-19 pandemic. Sex Transm Infect. 2022;98(Suppl 1):A48–A49. doi: 10.1136/SEXTRANS-BASHH-2022.99. [DOI] [Google Scholar]
- 23.French R.S., Geary R., Jones K., Glasier A., Mercer C.H., Datta J., et al. Where do women and men in Britain obtain contraception? Findings from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3) BMJ Sex Reprod Health. 2018;44(1):16–26. doi: 10.1136/jfprhc-2017-101728. [DOI] [PMC free article] [PubMed] [Google Scholar]